Today, when we talk about networking, we all pretty much use the TCP/IP model. It’s the standard. The gold standard. But it wasn’t always this simple.
Not too long ago, networking was like the Wild West — every vendor did their own thing, and if you had devices from different manufacturers, you needed a lot of patience (and a lot of technical know-how) to make them talk to each other.
Back in the 1970s and 1980s, IBM held a dominant market share and created its own proprietary network model called SNA (Systems Network Architecture). While IBM was busy claiming the throne, other vendors were off doing their own thing, building incompatible models. The result? A network engineer’s worst nightmare: a patchwork of models that didn’t always play well together.
-…
Today, when we talk about networking, we all pretty much use the TCP/IP model. It’s the standard. The gold standard. But it wasn’t always this simple.
Not too long ago, networking was like the Wild West — every vendor did their own thing, and if you had devices from different manufacturers, you needed a lot of patience (and a lot of technical know-how) to make them talk to each other.
Back in the 1970s and 1980s, IBM held a dominant market share and created its own proprietary network model called SNA (Systems Network Architecture). While IBM was busy claiming the throne, other vendors were off doing their own thing, building incompatible models. The result? A network engineer’s worst nightmare: a patchwork of models that didn’t always play well together.
The Need for a Universal Model
So, what did network engineers do when faced with these different models? Well, they had to work their magic. They spent hours trying to make these models exchange information—essentially translating languages between incompatible systems. It was like having two people speak different languages, trying to have a conversation with a dictionary in hand.
Let’s take a look at this visual from the early days:
Enter the ISO and OSI Model
To solve this mess, the ISO (International Organization for Standardization) stepped in. They launched a formal effort, bringing together technologically advanced countries to create a standardized networking model. This led to the creation of the OSI (Open Systems Interconnection) model, a seven-layer framework designed to make devices from different vendors talk to each other.
The OSI model, with its structure and defined layers, gave engineers a formal way of understanding how networks should function. It was an ideal concept, but it faced a bit of a challenge — it was complex and slow to catch on.
The Rise of TCP/IP
Meanwhile, across the pond, the US Department of Defense (DoD) had a more practical (and less formal) approach to creating a networking standard. They needed something that could scale across different systems and be reliable in critical situations. Enter TCP/IP.
This model wasn’t born out of a bureaucratic committee but rather from university volunteers who saw the need for a simpler, more adaptable networking system. Over time, TCP/IP became the dominant model, outpacing the OSI model and taking the world by storm.
The Evolution of Networking Models
For a while, both the OSI and TCP/IP models coexisted in organizations. But, like all good things, OSI eventually took a backseat. The simplicity and practicality of TCP/IP won out, and it remains the foundational networking model we use today.
Now that you’ve got a taste of how the TCP/IP model came to be, it’s easy to see how far networking has evolved. From the SNA model to the rise of OSI, and finally the triumphant TCP/IP, it’s been a journey of collaboration, frustration, and ultimately standardization.